The Time I Assumed Too Much About User Needs — And How AI Helped Me Listen Better
Building AI-powered tools for everyday life in Kenya often starts with an assumption — one that can be wrong. This is the story of when I assumed too much about user needs and how AI collaboration helped me pivot and deliver something that actually worked.
"The Time I Assumed Too Much About User Needs — And How AI Helped Me Listen Better Building AI-powered tools for everyday life in Kenya often starts with an assumption — one that can be wrong. This is the story of when I assumed too much about user needs and how AI collaboration helped me pivot and deliver something that actually worked. Every product I've built has started with an assumption — and more than once, those assumptions were wrong. One of the most valuable lessons I've learned came from a project that didn't go as planned. It was during the early days of Local Dialect, the language learning app focused on Kenyan and African languages. I assumed that users would want a highly structured, grammar-heavy approach — much like the tools available for learning English or Swahili. I was wrong. The Assumption That Didn't Fit I believed that the way to teach a language was through grammar, vocabulary, and sentence construction. I designed the first version of Local Dialect with that in mind. But when I started testing with users, I found that many of them preferred a more conversational, context-driven approach. They wanted to learn how to say things in real-life situations, not just memorize rules. My initial design didn't reflect that. It was a disconnect that I didn't see coming. How AI Helped Me See the Gap Working with AI collaborators like Claude and Codex, I started to re-examine the user flow and content structure. They helped me realize that the traditional grammar-based model wasn't the only way to go. Instead, I could build a learning experience that was more like a conversation — one that used real-life scenarios and dialogue. This was a shift in mindset that I wouldn't have made on my own. The AI didn't just help me code; it helped me understand my users better. Pivoting and Learning from the Mistake The mistake of assuming too much about user needs was a turning point for Local Dialect. I rebuilt the app around a more conversational format, and the response was immediate. Users engaged more, and the learning process felt more natural. I learned that assumptions — even well-intentioned ones — can lead us astray. But when we're open to listening, even through AI, we can find better solutions. What Changed as a Result That experience taught me the value of humility in product development. It also showed me how AI can be a partner in understanding users, not just in coding or content generation. I've since applied this lesson to other projects — like Arise & Shine Transporters, where I've had to listen closely to the needs of transporters and avoid making assumptions about how they work. The result is better products, more engaged users, and a deeper understanding of what real people need. You're Not Alone in This Mistake If you're reading this, you're probably someone who's made a mistake — maybe you assumed too much about a user's needs, or you tried something that didn't work. I want you to know that you're not alone. Every product I've built has had its share of missteps. But what matters is what we learned from them. And sometimes, that learning comes from AI — not just as a tool, but as a
